Maternal morbidity and mortality associated with interpregnancy interval: cross sectional studyBMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7271.1255 (Published 18 November 2000) Cite this as: BMJ 2000;321:1255
All rapid responses
We are gratified by the interest in our study expressed by Linnane et
It is worthwhile to mention that all their concerns regarding the
methodology of our research were discussed by us in the paper.
since less than 2% of all Latin American births are represented by our
database, our results may not be generalised to the whole of the Latin
American population and to the developed world. However, when we
replicated the entire analyses by country, the effects of interpregnancy
interval on maternal morbidity were essentially unchanged. The few
maternal deaths did not allow us to analyse by country.
Secondly, of the
520 689 parous women recorded in our database, we excluded 63 800 (12.3%)
for whom information on interpregnancy interval or adverse maternal
outcomes was missing or implausible. As stated in our paper, the accuracy
of specific diagnoses registered in our database has not been extensively
checked and only local medical record verifications were done. Therefore,
our data are limited to a certain extent. However, overall rates of
adverse maternal outcomes in this data set were similar to those reported
in other studies, which would add support to the accuracy of diagnoses.
Thirdly, in the present study we were unable to evaluate socio-economic
factors other than maternal education and cohabitation of parents because
these data were not available from the database. We agree that future
studies could consider other socio-economic factors such as family income
and race. However, it is not easy to include a variable relating to
ethnic origin due to the great mixture of races in Latin America and the
Caribbean. With regard to religious persuasion, it is doubtful that this
variable is related to socio-economic condition in Latin America and the
Fourthly, we agree that the relation between long
interpregnancy interval and preeclampsia-eclampsia could be confounded by
change of partner. However, we have controlled for the influence of other
possible confounding factors of this relation, such as older age, history
of chronic hypertension, and elevated body mass index before pregnancy,
and lack of previous miscarriage and smoking. The variable change in
paternity was not available to us for analysis.
Thanks again for their interest in our work.
Agustín Conde-Agudelo and José M. Belizán
Competing interests: No competing interests
Editor--Conde-Agudelo et al1 have examined the effect of pregnancy
interval on maternal morbidity and mortality. Specifically they looked at
456,889 births over a twelve-year period as recorded in a database that
registers half a million births a year. We welcome their study; however,
their methodology raises some concerns.
Firstly, the study is not population based but rather hospital based and
therefore represents less than 2% of all births in the Latin American and
Caribbean region. As a result, they have probably excluded all isolated or
community births which tend to occur more frequently among vulnerable
mothers anyway. Therefore, how representative is their sample of the whole
population? This would threaten the generaliseability of their findings
both to the developed world and to the rest of Latin American and the
Secondly there is no detail about the CAVEAT’s that should be applied to
data from databases and surveillance systems. (C = completeness, A =
accurate, V = valid, E = evaluated, A = available, T = timely.)
Thirdly, the authors appear not to have considered socio-economic factors
other than maternal education and cohabitation of parents, which by
themselves may not be the most appropriate markers. Given the vastly
disparate cultural areas they covered, should a variable relating to
ethnic origin or religious persuasion also be included to be more
meaningful here and for use in other studies?
The authors explained an increased risk of eclampsia and pre-eclampsia
amongst women with an interpregnancy interval of five years or more by
suggesting that a protective effect from an earlier pregnancy wanes after
five years. However, more conventional theories of the pathogenesis of pre
-eclampsia suggest that maternal exposure to a paternal antigen is
responsible.2 3 These more conventional hypotheses suggests that second
pregnancies are less likely to result in pre-eclampsia if the first one
was event free. This paper does not contradict this hypothesis, since long
interpregnancy intervals may actually be associated with the acquisition
of new reproductive partners. The findings on long interpregnancy interval
would be of much greater interest had partner data been included.
Finally, the authors concluded by recommending family planning to increase
interpregnancy intervals and thus avoid the risks of short intervals.
While this is a valid suggestion the opportunity to compare short
interpregnancy interval risks with those of contraceptive is needed.
Competing interests:None declared.
Eithne Linnane, Specialist Registrar in Public Health Medicine,
Philip Watson, Specialist Registrar in Public Health Medicine,
Bro Taf Health Authority,
Temple of Peace and Health,
1 Conde-Agudelo A, Belizán JM. Maternal morbidity and mortality
associated with interpregnancy interval :cross sectional study. BMJ
2 Trupin LS,Simon LP,Eskenazi B. Change in paternity:a risk factor for
preeclampsia in multiparas. Epidemiology1996;7:240-4.
3 Dekker GA, Robillard PY, Hulsey TC. Immune maladaptation in the etiology
of preeclampsia: a review of corroborative epidemiologic studies.
Obstetrical and Gynaecological survey. 1998;53(6):377-82.
Competing interests: No competing interests